Aquatic Plant Control Research Program Salvinia molesta D ...

us Army Corps of Engineers© Engineer Research and Development Center

Aquatic Plant Control Research Program

Salvinia molesta D. S. Mitchell (Giant Salvinia) in the United States: A Review of Species Ecology and Approaches to Management D. G. McFarland, L. S. Nelson, M. J. Grodowitz, June 2004 R. M. Smart, and C. S. Owens

200W730 056 Approved for public release; distribution is unlimited.

Aquatic Plant Control Research Program ERDC/EL SR-04-2 June 2004

Salvinia molests D. S. Mitchell (Giant Salvinia) in the United States: A Review of Species Ecology and Approaches to Management D. G. McFarland, L S. Nelson, M. J. Grodowitz

Environmental Laboratory U.S. Army Engineer Research and Development Center 3909 Halls Ferry Road Vicksburg, MS 39180-6199

R. M. Smart, C. S. Owens

Lewisville Aquatic Ecosystem Researcli Facility U.S. Army Engineer Research and Development Center 201 E. Jones Street Lewisville, TX 75057

Final report

Approved for public release; distribution is unlimited

Prepared for U.S. Army Corps of Engineers Washington, DC 20314-1000

ABSTRACT: Over the past 70 years, the free-floating aquatic fem Salvinia molesta D. S. Mitchell (giant salvinia) has spread from its native range in Brazil to many tropical and subtropical regions. Though innocuous within its native range, elsewhere this species is an aggressive menace that has had devastating ecological and socioeconomic impacts on aquatic systems in parts of Africa, Sri Lanka, India, Australia, New Guinea, and the Philippines. In the United States, the plant is estebHshed in waterways in at least 10 states (mainly in the south) and is expected to continue to expand in areas generally where Eichhomia crassipes (Mart.) Solms (water hyacinth) persists. Listed as a Federal Noxious Weed since 1984, S. molesta is prohibited from importation to the United States and from transport across state lines. Dense mats of S. molesta can suppress growth of native vegetation and degrade water qualily, fish and wildlife habitat, and numerous other ecological values. Notably, massive infestations have occurred in the S%vinney Mareh Complex, Texas, in the Lower Colorado River, Arizona/California, and in Lake Wilson and Enchanted Lake, Hawaii.

This report presente a review of available information on the growth, distribution, and ecology of S. molesta. Information is provided on tiie plant's taxonomic status, its field characteristics, phenology, and spread overeeas and in the United States. Growth responses of S. molesta in relation to environmental variables (e.g., temperature, nutrients, light, pH, conductivitj') are emphasized as are impacts of the species on the environment and other aquatic oiganisms. Difierent technologies (i.e., physical, chemical, biological, and integrated) applied to control S. molesta infestations are discussed along with information on the effectiveness of these procedures and their need for further study.

DISCLAIMER: The contents of this report are imt to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endoreement or approval of tlie use of such commercial products. All product names and trademaiks cited are the propertj' of their respective owneis. The findings of this report arc not to be construed as an official Department of the Am^' position unless so designated by other authorized documents.

Contents

Preface v

1—Introduction 1

Background 1 Report Objective 1

2—Plant Description and Ecology 3

Taxonomic Status 3 Geographical Distribution 5

Native and overseas distribution 5 History of spread in the United States 5

Field Recognition 7 Reproduction 8 Dispersal 8 Productivity..... 9 Requirements for Growth 10

Temperature 10 Light 10 pH 11 Conductivity 11 Salinity 11 Nutrients 11

Impacts on the Envirormient 12 Detrimental 12 Beneficial 14

3—Management Options 15

Chemical Control 15 Background — Herbicide investigation and use overseas 15 Herbicide investigation and use in the United States 16

Biological Control 18 Background 18 Host-specificity testing 19 Description 21 Biology 21 Feeding damage 21 Plant impacts 21 U.S. operational status 21

III

Physical Control Conventional procedures for floating aquatic plants. Direct removal Barriere Habitat alteration.....

Ecosystem Approach Prevention.... Eradication Integrated pest management "Filling the niche" Nutrient management

References

.....22 22 22 23 23 24 24 24 .....24 25 25

.....26

SF298

Preface

The review presented herein was performed as part of the Aquatic Plant Control Research Program (APCRP), U.S. Army Engineer Research and Development Center (ERDC), Vicksburg, MS. The APCRP is sponsored by Headquarters, U.S. Army Corps of Engineers (HQUSACE), and is assigned to the ERDC under the purview of the Environmental Laboratory (EL). Funding for publication was provided under Department of the Army Appropriation No. 96X3122, Construction General. Mr. Robert C. Gunkel, Jr., EL, was Program Manager for the APCRP. Program Monitor during this research was Mr. Timothy R. Toplisek, HQUSACE.

This manuscript was prepared by Ms. Dwilette G. McFarland, Dr. R. Michael Smart, Dr. Michael J. Grodowitz, and Ms. Chetta S. Owens, Aquatic Ecology and Invasive Species Branch, Ecosystem Evaluation and Engineering Division (EEED), and Dr. Linda S. Nelson, Environmental Processes Branch, Environmental Processes and Engineering Division (EPED), EL. Both Dr. Smart and Ms. Owens are researchers at the ERDC, Lewisville Aquatic Ecosystem Research Facility (LAERF), Lewisville, TX. The review was performed under the general supervision of Dr. Elizabeth Fleming, Acting Director, EL, and under the direct supervision of Dr. Al Cofirancesco, Chief, EEED, EL.

The authors gratefully acknowledge librarians of the Information Technology Laboratory, ERDC, with special thanks to Messrs. Donald Kirby and Paul Taccarino, and Ms. Jimmie Perry, Information Management Reference Library (IM-R), for locating and obtaining the requested reference materials. Apprecia- tion is also extended to Dr. Judy F. Shearer and Ms. Dian H. Smith, EL, ERDC, for their helpful comments and suggestions in review of an earlier version of this manuscript.

COL James R. Rowan, EN, was Commander and Executive Director of ERDC. Dr. James R. Houston was Director.

1 Introduction

Background

Salvinia molesta D. S. Mitchell (giant salvinia), a free-floating aquatic fern, is one of the world's worst aquatic weeds, second only to Eichhornia crassipes (Mart.) Solms (water hyacinth) (Holm et al. 1977; Barrett 1989). Though much smaller than E. crassipes, S. molesta has a faster growth rate that, under optimal conditions, enables it to double in number and biomass in less than 3 days (Farrell 1978, 1979; Harley and Mitchell 1981; Barrett 1989). Problems caused by S. molesta result from rapid vegetative reproduction in which rhizomatous colonies are produced and disseminated through fragmentation (Room 1986a, 1990). Excessive growth of this species prevents light penetration of the water column by forming dense floating mats that can shade out favorable vegetation and degrade habitat for fish and wildlife. Dense mats oiS. molesta can impede water-based transport and recreation, reduce water quality and aesthetic values, and clog irrigation and power generation intakes (Thomas and Room 1986a; Room et al. 1989). Public health problems have also been associated with nuisance growth of the plant, which restricts access to potable water supplies and harbors mosquitoes and other vectors of human diseases (Beimett 1975; Room etal. 1989).

Over the past 70 years, S. molesta has spread from its native range in South America to tropical and subtropical regions around the world (Oliver 1993; Jacono and Pitman 2001). Its explosive growth has had devastating socioeconomic impacts in parts of Africa, Sri Lanka, New Guinea, the Phihppines, and Australia (Oliver 1993; Chilton et al. 2002). Lake Kariba on the Zambesi River (Africa) experienced one of the most phenomenal invasions when in 1962 at its peak the weed covered a quarter of what was then the world's largest reservoir (Barrett 1989). In the United States, severe infestations have occurred in the Swinney Marsh Complex, Texas, in the Lower Colorado River, Arizona/ California, and in Lake Wilson and Enchanted Lake, Hawaii (Jacono and Pitman 2001; TenBruggencate 2003).

Report Objective

The invasion of many U.S. waterways by S. molesta in recent yews has increased the demand for information on the ecology and management of this weed. To improve programs to control S. molesta and prevent its rapid spread.

Chapter 1 Introduction

knowledge is needed of current management practices, their levels of efficiency, and vulnerabilities of the species in different conditions.

The objective of the present report is to review available information on the gro-wth, distribution, and management of S. molesta. Information is provided on the plant's taxonomic status, field characteristics, life history, and occurrence overeeas and in the United States. Ecological information on S. molesta is also included, emphasizing environmental tolerances, gro-wth requirements, and impacts on the environment. Efforts toward managing this nuisance species using different control technologies (e.g., physical, chemical, and biological) are discussed with an indication of degree of efficacy, operational status, and/or need for further investigation.

Chapter 1 Introduction

2 Plant Description and Ecology

Taxonomic Status

Kingdom: Plantae; Subkingdom: Tracheobionta; Division: Pteridophyta; Class: Filicopsida; Order: Hydropteridales; Family: Salviniaceae; Genus: Salvinia Seguier; Species: Salvinia molesta Mitchell (Plants National Database 2003).

Salvinia molesta D. S. Mitchell belongs to a monogeneric family (Salviniaceae) of free-floating aquatic ferns coated with velvety hairs on the leaf surfaces (Figure 1). Up to 12 species of the genus {Salvinia) have been reported worldwide {S. auriculata Aubl. (= S. rotundifolia Willd.), S. biloba Raddi, S. cucullata Roxb. ex Bory, S. hastata Desv., S. herzogii de la Sota, S. martynii Spruce, S. molesta D. S. Mitch., S. natans (L.) All., S. nymphellula Desv., S. oblongifolia Mart., S. minima Baker, S. sprucei Kuhn], seven of which {S. oblongifolia, S. sprucei, S. minima, S. auriculata, S. herzogii, S. biloba, andS. molesta) originated in the neotropics (Mitchell and Thomas 1972; Sculthorpe 1985; Jacono and Pitman 2001). Salvinia molesta along with three other closely related species, i.e., S auriculata, S. herzogii, and S. biloba, comprise the taxonomic assemblage known as "S. auriculata complex" (Fomo 1983). Members of this complex exhibit rows of cylindrical hairs with branches joined at the tips to form an "eggbeater or cage-like" structure. All four species are listed on the Federal Noxious Weed List which prohibits their introduction to the United States as well their interstate transportation. At present, S. molesta is the only one of these four species that occurs in the United States outside of cultivation (Jacono 2003a).

Initially identified as a form of 5'. auriculata, S. molesta was reclassified in 1972 based on details of the male sporocarps or fruiting bodies (Mitchell 1972). Current taxonomic treatments noting the confiision over species identity indicate S. auriculata auct. non Aubl. (misapplied) as a probable synonym (Global Compendium of Weeds (GCW) 2004). Salvinia molesta is also known by a variety of common names (e.g., salvinia, giant salvinia, African pyle or payal, Kariba weed, aquarium watermoss, Australian azolla, water fern, water spangles, and giant azolla) generally recalling plant size, specific infestations, and aquatic nature of the species (Oliver 1993; Hassler and Swale 2002).

Chapter 2 Plant Description and Ecology

Sahinia molesta Giant salvinia

Coi^^t 20<M lMveB% of Ftod* Cents- for Aquatic and fcyasive PJants

Figure 1. Morphology of Salvinia molesta D. S. Mitchell: a. Individual plant, showing sporocarps, floating leaves (i.e., fronds), and rootlike submerged leaf; b. Floating leaf, cross section; c. Primary form; d. Secondary form; e. Tertiary form (permission to reprint granted by University of Florida, Center for Aquatic and Invasive Plants 2000)

Chapter 2 Plant Description and Erology

Geographical Distribution

Native and overseas distribution

Salvinia molesta is considered native to southeastern Brazil, in a subtropical zone (between latitudes 24° 05' S and 32° 05' S), extending inland to an elevation of about 900 m (Fomo and Harley 1979; Fomo 1983). It was first observed outside its native range in Sri Lanka in 1939 and has since become a serious nuisance in over 20 countries (Room 1986a; Room et al. 1989; Room 1990; Oliver 1993; Storrs and JuUen 1996). Beyond the United States, S. molesta is presently established in Australia, New Zealand, Fiji, the Philippines, India, Indonesia, Malaysia, Singapore, and Papau, New Guinea. It also plagues aquatic systems in Afiica (the Ivory Republic, Ghana, Zambia, Kenya, Namibia, Botswana, South Afiica, and Madagascar), South America (Columbia and Guyana), and two Caribbean countries (Cuba and Trinidad; cf. syntheses by Oliver 1993; Storrs and Julien 1996). Historically notable infestations have occurred in the Sepik River of Papua New Guinea, and in Afiica in the Zambezi River, Lake Naivasha, Kariba Lake, and the Chobe River System. Most recently, S. molesta was reported in southern Kalimantan (formerly Borneo), where rivers, swamps, and rice paddies are becoming increasingly overgrown (Jacono and Pitman 2001).

History of spread in the United States

For nearly 2 decades, S. molesta has been cultivated in the United States (Figure 2) as an ornamental plant provided by the horticulture industry (Harley and Mitchell 1981; Nelson 1984; Jacono 2003a). The species was first reported outside cultivation in this country in 1995 in a private 0.6-ha (1.5-acre) pond in southeastern South Carolina (Johnson 1995; Jacono and Richerson 2003). That same year, the plant was removed with chemical herbicide treatment and no fiarther cases in South Carolina have been reported since then (Chilton et al. 2002; Jacono and Richerson 2003). However, in 1998, new outbreaks were discovered in Texas and Louisiana, initially in small ponds and later in Toledo Bend Reservoir, oxbow lakes of the Sabine River, and swamplands of Swinney Marsh (Chilton et al. 2002; Jacono and Pitman 2001; Jacono and Richerson 2003). The infestation of Toledo Bend Reservoir was the first in pubUc U.S. waters and posed a significant threat for interstate dispersal because of the reservoir's large size, i.e., approximately 75,272 ha (186,000 acres), and location on the state line between Texas and Louisiana (Chilton et al. 2002). A primary concem was the reservoir's use by many thousands of sportsmen who could inadvertently transport the weed on their boating equipment across these two states and elsewhere. During the year 2000 growing season, 485.6 ha (1,200 acres) of Toledo Bend were chemically treated, but these efforts appear to have been only marginally successfiil (Jacono and Pitman 2001; Chilton et al. 2002).


Hawaiian Islands

ÛSGS

SaMnis mofes^

: Sates ssShffiôJlfeônês

July 2003

Figure 2. Recorded distribution of Salvinia molesta in the United States

By the end of 1999, S. molesta had invaded over 50 localities in southern tier states (Texas, South Carolina, Louisiana, Georgia, Florida, Alabama, Mississippi, Arizona, California) and Hawaii (Jacono 2003a). A rapidly expanding infestation was documented in April of that year, when S. molesta in Enchanted Lake, Kailua, Hawaii, threatened the habitat of three endangered waterbird species, i.e., the Hawaiian coot (Fulica alai), Hawaiian gallinule {Gallinula chloropus sandivicensis), and Hawaiian stilt {Himantopus mexicanus knudseni) (Jacono and Richereon 2003). The following August, major infestations were discovered in the Lower Colorado River, at the Imperial National Wildlife Refuge bordering Arizona and California. Plants from the Palo Verde Irrigation District apparently initiated colonies that have since penetrated the Mexican border via the Colorado River (Jacono and Pitman 2001). Many smaller infestations were chronicled in 1999 for various aquatic systems in Scale and Auburn, AL; Moselle, MS; Houma, LA; Houston (vicinitj'), Lovelady, Freeport, Alvin, Mont Belvieu, and Flower Mound, TX; Atlanta (vicinity), GA; Oahu, HI; and Naples, PL (see Jacono and Richereon 2003 for details).

The year 2000 marked the firet sightings of S. molesta in North Carolina in low-lying areas near Burgaw and in Jacksonville, and in ponds in Wilmington on the Cape Fear Peninsula (Kay 2002; Jacono and Richereon 2003). New reports of the species have since become less frequent, though many existing infestations have remained problematic. Today, the most widespread infestations occur in Texas and Louisiana where 4 public reservoirs, 7 rivere and streams, 2 large marehlands, and over 25 ponds have been impacted (Jacono and Pitman 2001). Swinney Mareh Complex on the Lower Trinity River, in Liberty County, Texas, continues to support one of the most serious infestations of S. molesta in the

Chapter 2 Plant Description and Eojlogy

United States (Jacono 2002; Jacono and Richerson 2003). Accounts by the U.S. Geological Survey (USGS) now show S. molesta naturalized in 10 states (North Carolina, Georgia, Florida, Alabama, Mississippi, Louisiana, Texas, Arizona, California, Hawaii), with South Carolina being the only state where the plant has been eradicated (Chilton et al. 2002; Jacono 2003a; Jacono and Richerson 2003).

Field Recognition

Salvinia molesta forms free-floating colonies of potentially independent modules (ramets), each held together by a horizontal stem (rhizome) just below the water surface (Figure 1). The rhizome of an intact module bears three distinct leaves (or fronds), along with four or five buds that are relatively inconspicuous. Two leaves are green, emergent or floating, and ovate to oblong, while the third (a modified submerged leaf) is brown, highly dissected, and hangs underwater. The upper surface of each floating leaf has a prominent midrib and is covered by rows of white, bristly hairs (trichomes), topped with four branches united distally to form a structure resembling an "eggbeater" or "kitchen whisk." These hairs give the leaves a velvety appearance and serve as air traps to repel water and aid in flotation (Harper 1986). The submerged leaf resembles "root mass" and can elongate to create resistance to the water and thus helps to stabilize the plant. Sori or sporocarps develop in long chains among filaments of the submerged leaves, and in mature plants, are formed in large quantities but are functionally sterile (Loyal and Grewal 1966). Bonnet (1955), Croxdale (1978, 1979, and 1981), and Harley and Mitchell (1981) have provided detailed descriptions of the anatomy and development of .S. molesta ramets.

Three main phenotypes or growth forms of .S. molesta have been recognized with a continuum of intermediate morphologies (Figure 1):

a. The primary {or primary-invading) stage occurs in isolated or widely spaced plants, during the initial invading stage of the infestation. The plants produce small, dehcate, oval leaves, from (2) 10 to 15 mm wide, that lie flat upon the surface of the water. This growth form may also be observed in plants recovering from damage or in uncrowded conditions in shade or rich nutrient culture (Harley and Mitchell 1981).

b. The secondary (or open water-colonizing) stage is evident in plants that have grown in open water for some time, either freely or on the edge of stable mats. Stem intemodes are longer and larger than in the primary stage; leaves are slightly cupped but do not overlap, and the entire lower leaf surface is in contact with the water. Sizes of the leaves vary widely from about 20 to 50 mm in diameter.

c. The tertiary (or mat) stage occurs under crowded conditions typically associated with a mature infestation. This form is a relatively robust with short intemodes; the leaves are large (up to 60 mm in diameter), heart- shaped or oblong, and deeply keeled. As crowding increases, the leaves are pushed upward to erect and may become packed into mats up to 1 m thick.


Ashton and Mitchell (1989) have described a survival form that occurs in S. molesta populations growing in hareh environments, e.g., under poor nutrient conditions. In this form, the plant grows slowly, bearing up to five pairs of leaves that are flat, sometimes yellowsh, and typically small (5 to 8 mm in diameter).

Reproduction

Salvinia molesta produces sporocaips but spores are rarely found, and when present, the spores are generally deformed and nonviable. Because the plant is pentaploidal (2n = 45), and thus genetically unable to produce fertile spores, reproduction in S. molesta is believed to be solely asexual (Loyal and Grewal 1966). Rhizomes of the plant can break very easily and dmighter plante arise from buds that number up to five per node. In calm waters, fragmentation occurs when older ramets senesce, causing the plant to break, as a result of deterioration of the rhizome, into two or more daughter plants (Room 1983). Fragmentation may also result from mechanical damage from human ^tivities (e.g., harvesting and boating) or from friction between plants being moved by winds and water currents (Harley and Mitchell 1981; Room 1983,1990).

Dispei^al

Vegetative spread by fiagmentation is thought to be the major means of intra- and inter-lake dispersal of S. molesta. High mobility of the plant is fecili- tated by aerenchjTna tissue that increases buoyancy and enables movement over vast areas, particularly during flooding (Barrett 1989; Oliver 1993). Rises in water level during flooding can loosen and break up S. molesta mats, allowing wind and water movement to disperse parts of the mat and plant fragments (Barrett 1989). Cblonies can expand laterally over small distances by rhizome and bud growth, while fragments appear to be the dominant means of long- distance dispersal. Fragments may spread, for example, when parts of the plant adhere to fishing equipment, boats, trailere, or other vehicles (Chilton et al. 2002). Animals may also contribute to vegetative spread: hippos in Afiica and water buffalo in Australia reportedly have carried S. molesta over short distances (Miller and Wilson 1989; Storre and Julien 1996). Raccoons, wildfowl, turtles, and other wetland animals could possibly serve as vectois, but their role in spreading S. molesta h^ not been investigated extensively.

The attraction of S. molesta as an ornamental plant has led to its interconti- nental transport via aquarium and landscaping trades. Its introduction to North America, Asia, Afiica, and other continents has been linked to cultivation ^tivities of botanical gardens and commercial horticulture sites (Harley and Mitehell 1981; Nelson 1984; Thomas and Room 1986a; Oliver 1993; Jacono 2003a). Plants initially introduced to the United States probably arrived in Florida from Sri Lanka, as cargo for direct sale or as a contaminant in an aquatic plant ship- ment (Nelson 1984; Oliver 1993; Jacono 2003a). A long-standing assumption has been that naturalized populations of exotic plants occur by these plants being dumped into nearby waterways or being "seeded" deliberately in the wild for fiiture marketing.


Jacono (2003a) has reported S. molesta in cultivation in public and private aquatic gardens and nurseries in 17 states in the United States (Figure 2). These sites are potential sources of release into natural systems, and interestingly, many are located in areas where S. molesta infestations have been documented. To date, among the 11 states with records of S. molesta in nature, 8 states contain sites where the species is in cultivation (Jacono 2003a).

Sahnnia molesta is prohibited, as a Federal Noxious Weed, from transport across state lines and from being imported to the United States (Chilton et al. 2002; Jacono and Pitman 2001; Jacono 2003a). However, for the species to be restricted from sale, cultivation, and ownership within a given state, the plant must be listed by the state as a State Noxious Weed (Jacono 2003a). Presently, S. molesta can be freely cultivated and sold within 42 states (including Hawaii) as long as it is not transported across state lines (Jacono 2003a). It is hsted as a State Noxious Weed in Arizona, California, Florida, Georgia, Louisiana, North Carolina, South Carolina, and Texas but should be prohibited by others, especially those states with a history of infestation within their boundaries (e.g., Mississippi, Alabama, and Hawaii) (Jacono 2003a).

Productivity

Rates of growth and reproduction are formidably high in S. molesta. In the laboratory, its leaf doubling time ranged from ~ 2 to 4 days (Gaudet 1973; Mitchell and Tur 1975; Mitchell 1979a), and from ~ 1 to 8 days under field conditions (Finlayson 1984; Room 1986a). Rates of leaf production strongly reflect changes in season. In midsummer, the number of leaves were found to double in 2.2 to 2.7 days, and in 40 to 60 days in winter in Lake Moondarra, Mt. Isa, Australia (Farrell 1979; Harley and Mitchell 1981; Finlayson 1984). Mitchell and Tur (1975) reported that in Lake Kariba, Zimbabwe, Africa, doubling time (in terms of leaf number) was reduced from 14.3 days in winter to 8.1 days in summer. Averaged across all seasons, doubling time varied with growth form - being shorter in S. molesta in the secondary (8.6 days) than in the tertiary form (11.6 days).

Rapid growth rates of 5'. molesta enable it to cover still and slow-moving waters with dense surface mats, up to 1 m thick, depending on mat age and degree of compaction (Harley and Mitchell 1981; Thomas and Room 1986a). In Lake Kariba, the number of plants was as high as 4,672/m^ in stable mats and ranged between 200 to > 10,000/km^ (mean = 3,290) in plants drifting in open water (Mitchell 1970). Live biomass in the mats was found to vary considerably, ranging from 250 to 600 g/m^ dry wt, which approaches the 670 to 1,620 g/m^ dry wt observed for £■. crassipes (Mitchell 1979a). Some of the highest rates of doubling in percent cover (1.3 days), leaf number (1.4 days), and fresh weight (1.8 days) have been recorded for S. molesta in a sewage lagoon near Lake Moondarra, with no evidence of toxicity (Finlayson 1984). Mitchell and Tur (1975) estimated that a mat growing at a conservative rate of 5 percent per day (doubling about every 14 days) would produce ~ 45.6 to 109.5 tons/ha/yr. However, since this estimate was based on the lower end of the range of growlh rates in Lake Kariba, they speculated that even greater production in S. molesta might result under more favorable conditions.


Requirements for Growth

Temperature

Results of numerous laboratory and field studies (Gaudet 1973; Rani and Bhambie 1983; Caiy and Weerts 1983a, 1983b; 1984; Finlayson 1984; Room 1986b) have enabled the development of a predictive model for the growth of S. molesta in relation to temperature (Room 1988). Accoiding to the model, grov^th of 5. molesta follows a bell-shaped curve, increasing to a maximum near 30 °C, with upper and lower limits of about 40 and 5 °C. More recently, Whiteman and Room (1991) showed that temperatures < -3 °C or > 43 °C are lethal to buds in exposures > 2 hr. Their findings are consistent with temperature regimes within the plant's distribution, which extends from the equator to regions that experience frost but not the formation of ice on the water. While freezing temperatures could potentially kill the most exposed plants, the population could recover and persist from xmderlying protected plants. Whiteman and Room (1991) concluded that near its limits in hot and cold climates, S. molesta is more likely to survive in large bodies of water where thermal capacity dampens temperature fluctuation.

Based on results of the above studies, the U.S. Department of Agriculture (USDA) has formulated an "expected range" for the expansion of S. molesta within the United States. TTiis range, as reported by Jacono (2003b), includes the Atlantic coastol plain, from southeastem Virginia to southern Florida, the Gulf coast states, north to central Califomia, and southern Arizona. These areas generally coincide with regions in Zone 8 of the USDA Plant Hardiness Map (U.S. National Arboretum 2003). Emergent growth of the plant dies back in winter in North Carolina (Zone 8, at latitude 34.4° N), and in north Texas (Zone 7b, at latitude 33° N) but generally recovere to dense levels the following growing season (Jacono 2003b).

Light

As a floating aquatic plant species with access to light at the water surface, S. molesta in most cases may be less affected by light as compared with other environmental fectors, e.g., temperature and nutrients. This may explain in part why so few studies have been conducted to assess growth of S. molesta in relation to light conditions. Current findings indicate that growth of this species is light saturated at ~ 4,500 kcal/mVday then is inhibited with further increase in light intensity (Rani and Bhambie 1983). Under favorably high light conditions, its grornlh is increased with increases in temperature up to about 30 °C (Mitchell and Tur 1975; Rani and Bhambie 1983). Significant interactive effects of light and temperature are reflected in seasonal changes in the rates of growth and biomass production in S. molesta. For example, outdoor studies of this species showed tiiat during August to September, growth rates were 0.06 to 0.07 g/g/day and doubling time was 9.8 days in full sun and 11.8 days in shade. How^ever, under cooler temperatures that existed from December to Februaiy, growth rates were 0.01 to 0.02 g/g/day, and doubling time w^ 23 days under both light conditions (Rani and Bhambie 1983).

10 Chapter 2 Plant Description and Ec»logy

pH

Cary and Weerts (1984) found that over a range in pH from 5 to 8, S. molesta produced the greatest amount of biomass at pH 6. Total biomass at pH 6 was 11, 54, and 59 percent greater than at pH 5, 7, and 8, respectively. The predilection of 5'. molesta for slightly acidic or near neutral pH is supported by accounts that it grows well in nature at pH 6.0 to 7.4 (Johnson 1967; Gaudet 1973; Holm et al. 1977; Mitchell et al. 1980). Recent comparisons between two research ponds in Lewisville, TX, showed more extensive infestation of the pond with pH < 7.5 than the one with pH 8.5 or greater (Owens et al. in preparation). The ability of the plant to tolerate a wide range in pH is evidenced in its occurrence at pH 5.2 in Malaysia, pH 6.0 in Singapore, pH 7.4 in Zimbabwe, and pH 6.5-9.5 in other waters of Africa (Johnson 1967, as per Gaudet 1973; Holm et al. 1977; Storrs and Julien 1996).

Conductivity

Field surveys indicate S. molesta to infest waters with conductivities ranging from 239.3 (± 77.9) to 503.5 (± 446.2) ^S/cm (Room and Gill 1985). Sewage lagoons supporting dense growth of the plant had conductivities as high as 1,375.4 (±149.5) y&lcm (Room and Gill 1985). Although thick mats have been observed in waters with low conductivity (~ 100 (iS/cm), the leaves appeared chlorotic, very likely because of low availability of nitrogen (Mitchell et al. 1980). Small infestations can survive at about 2,000 ^S/cm (Storrs and Julien 1996), but higher conductivities (~ 4,800 |iS/cm) damage plant tissues and diminish likelihood of survival (Divakaran et al.l980).

Salinity

Salvinia molesta has a low tolerance for saline environments and does not colonize marine or brackish waters (Oliver 1993). This species can survive at salinities up to 20 percent that of seawater, with growth rate decreasing with increases in total dissolved salt concentration. Divakaran et al. (1980) demonstrated a 25-percent reduction in rate of growth oiS. molesta in water having a salinity of 3 ppt (or 10 percent that of sea water). Plants in solution with a salinity of 7 ppt had a doubling time of 108 days as opposed to 3.8 days for control plants (King and Mitchell unpubl. data, as per Harley and Mitchell 1981). Salinities > 7 ppt appear to be lethal, with death occurring at 11 ppt after 20-hr exposure; S. molesta died after only 30 min of exposure when cultured at 34 ppt, i.e., fiiU strength seawater (Divakaran et al. 1980; Harley and Mitchell 1981).

Nutrients

Since S. molesta is free-floating and is not rooted in sediment, nutrients required for growth must be obtained by portions of the plant in contact with the water column. Dissolved nutrients, especially nitrogen (N) and phosphorus (P), play a key role in determining morphological characteristics and rates of growth of iS. molesta colonies (Room and Thomas 1986a, 1986b). Concentrations of

Chapter 2 Plant Description and Ecology 11

nutrients in aquatic systems can fluctuate widely and, in heavily infested areas, are frequently below limits for accurate quantification (Storrs and Julien 1996). Therefore, measurements of nutrients in plant tissues, as compared to water samples, can provide a better assessment of nutrient availability for plant uptake overtime.

Previous studies of nutrient relationships have shown that S. molesta can accumulate high concentrations of nutrients to sustain growth when supplies become deficient (Gaudet 1973; Cary and Weerts 1984). Its gro^^lh ceased to be limited by N and P availability when tissue concentrations reached approximately 5 percent or more for N and 0.5 percent or more for P under laboratory' conditions (Caiy and Weerts 1983a, 1983b; Room 1986b). In > 1,700 samples of S. molesta collected from field sites in Australia and Papua New Guinea, tissue N concentrations ranged fi'om 0.62 to 4.0 percent and tissue F fi-om 0.03 to 1.07-percent drj' wt (Room and TTiomas 1986a, 1986b). The Australian average for concentrations of N in tissues of S. molesta was 1.59 percent, while tiiat for tissues from Papua New Guinea was 1.12 percent. Based on these low values, N was suggested to be the primary limiting nutrient, accounting for 40 and 80 percent of the variance in growth rates of S. molesta in the field (Room and Thomas 1986a, 1986b).

Salvinia molesta grows more rapidly when provided dissolved inoiganic N as mit rather than NO3" ions (Gaudet 1973; Cmy and Weerts 1983a). In solution with NH4-N or NH4NO3-N as a source of N, gro^^th of S. molesta was twice the amount achieved with equivalent quantities of NO3-N or urea-N (Gary and Weerts 1983a). Optimal levels of dissolved N depend on levels of other dissolved nutrients, especially P. High levels of biomass production and doubling times < 4 days occurred in S. molesta in solutions with N and P combinations varying fi-om 2 to 20 mg NH4-N/L and from 2 to 10 mg PO4-P/L (Gary and Weerts 1984).

Availability' of nutrients in the water is an important determinant of the survival strategy demonstrated by S. molesta colonies. In infertile water, S. molesta riiizomes are tough with few if any branches and fiagment when the oldest segments deteriorate (Mitchell and Tur 1975; Room 1983; Juhen and Bourne 1986). Survival is enhanced under infertile conditions by retaining senescing ramets so that the nutrients within them can be reallocated elsewhere in the colony (Room 1983). In contrast, S. molesta in fertile water has highly branched and brittle riiizomes that fragment sooner than in infertile water because of pressure from overcrowded ramets (Room 1983,1990). Senescing ramets are not retained probably because proliferation of dispereal units is selected under fertile conditions over retention of nutrients (Stons and Julien 1996).

Impacte on the Environment

Detrimental

By hindering the use of water resources, extensive mats of S. molesta threaten a multitude of environmental, economic, and human health interests. Dense growth of the plant forms a physical barrier on the water surface that prevents or impedes water use for recreational activities, such as swimming,

12 Chapter 2 Plant Description and Ecxilogy

boating, water skiing, and fishing (Holm et al. 1977; Barrett 1989; Chilton et al. 2002). Even when the mats are not fully impenetrable, the clusters of rhizomes and submerged leaves quickly become entangled around boat propellers, so that the watercraft may sooner or later be immobilized (Sculthorpe 1985). Wind and water movements can bank the plants into thick stabihzed mats that may accumulate in grids and sluices of dams and electrical generating installations, and plug up irrigation systems (Holm et al. 1977). Where water flow is restricted and rates of silt deposition increase, dredging operations may be needed to minimize potential adverse effects of flooding. In addition to causing the loss of crops and preventing agricultural operations, flooding is a prominent mechanism in breaking up the mats and distributing fi^agments that can produce new colonies (Sculthorpe 1985).

Occasionally, plants of various species will colonize thick S. molesta mats, forming "floating islands" of vegetation or mixed sudd communities (Sculthorpe 1985; Hohn et al. 1977). The colonizing species often include water pepper and knotweeds {Polygonum spp.), water primrose {Ludwigiapeploides, L. leptocarpa, and L. ascendens), grasses (Leersia hexandra, Hymenachne amplex- icaulis, and Panicum repens), sedges {Cyperus platystylis and C. cephalotus), bulrush {Scirpus cubensis) and even trees {Melaleuca spp.) (Holm et al. 1977; Mitchell et al. 1980; Storrs and JuUen 1996). Boughey (1963) documented more than 40 plant species that colonized S. molesta mats in Lake Kariba, central Africa. These deceptive islands have caused a number of livestock deaths by livestock sinking or breaking through the mats and drowning in deep water (Harper 1986).

Salvinia molesta can have other detrimental effects on the ecology of aquatic systems by restricting hght penetration and exchange of gases between the water and atmosphere. By curtailing the availability of light, 5'. molesta can outcompete many native species of submersed and floating plants, consequently reducing community diversity (Sculthorpe 1985). Water quality beneath the mats is almost always degraded by decreases in dissolved oxygen and pH, and increases in CO2 and H2S concentrations (Mitchell 1969). As the plants die, organic debris accum- ulates at the bottom of the water coliram and can threaten fisheries by creating a shallow-water environment less suited to fish breeding (Sculthorpe 1985). Fur- thermore, decomposition of the organic materials can greatly diminish dissolved oxygen supplies needed to support healthy fish populations and other biota (Hattingh 1961; Coates 1982; Oliver 1993). Rapid rates of nutrient uptake com- bined with relatively slow rates of decomposition enable S. molesta to tie up nutrients that could be used by other primary producers that contribute to complex food chains. The theoretical maximum rate of N uptake, calculated from rates of growth of .S. molesta, is about 8 mg N/g dry plant tissue/day or about 6,000 kg N/ha/yr (Room 1986b).

Heavy infestations oiS. molesta raise public health and economic concerns by serving as breeding habitat for vectors of human disease, and interfering with food production and market transport. The stagnant, shallow-water conditions and dense foliage produced by S. molesta mats are favored by many species of mosquitoes that transmit encephalitis, dengue fever, malaria (Creagh 1991/1992 as per Oliver 1993), and rural filariasis or elephantiasis (Room et al. 1989). In Lake Kariba (Zambia and Rhodesia), S. molesta mats fostered the buildup and

Chapter 2 Plant Description and Ecology 13

spread of Biomaphalaria boissyi, the snail that is the intermediate vector of bilharzia or schistosomiasis (Bennett 1975). In Sri Lanka, India, and Borneo, the plant is a serious pest in ricelands, while its encroachment into other Asian and African waterways h^ led to declines in tourism, hunting, and fish industries (Bennett 1975; Thomas and Room 1986a; Oliver 1993). Yet nowhere have the adveree effects of rapid grov^ih of S. moles ta been as dramatic as in flie flood- plain of the Sepik River, Papau New Guinea. There, entire villages, dependent on aquatic transportation, were abandoned because thick mats of S. molesta elimi- nated access to health care, food, maricets, and schools (Mitehell et al. 1980; Thomas and Room 1986a).

Beneficial

Despite limited possibilities, investigative attention has been directed toward finding value in the large amounts of biomass S. molesta produces. The plant has been used as a compost and mulch and as a supplement to fodder for livestock in some Asian countries (Oliver 1993). A few studies have examined its suitabilitj' in treating sewage effluent (Finlayson et al. 1982), papermaking, and the generation of biogas (Thomas and Room 1986a). However, none of these efforts has led to large-scale utilization, probably because high coste associated with labor and machinery. For such strategies 1x) be viable, a continued supply of S. molesta must be available, which could exacerbate problems for aquatic weed management (Thomas and Room 1986a).

14 Chapter 2 Plant Description and Ecology

3 Management Options

Chemical Control

Background — Herbicide investigation and use overseas

Numerous studies worldwide have evaluated and documented the use of chemical products to control S. molesta. According to Thomas and Room (1986a), the first attempts to chemically manage the plant were made in Sri Lanka in the 1940's, where high volume applications of emulsifiable oils containing pentachlorophenol (a chemical wood preservative) were used in rice paddies and waterways. In the years to follow (1960's to 1970's), a variety of chemicals were tested against S. molesta including: monuron (a herbicide), anhydrous ammonia, formalin (a fish parasiticide), niclosomide (a molluscicide), and thiram (a fimgicide) (Thomas and Room 1986a). While some of these products showed herbicidal activity against S. molesta in laboratory trials, most were never used on an operational scale.

Kam-Wing and Furtado (1977) assessed the effectiveness of the herbicides paraquat, diquat, nitrophen (as the coded compound, TOK E-25), and dalapon for control of 5*. molesta in Malaysia. Of these products, only paraquat and diquat showed potential in laboratory studies. Paraquat applied as a fohar spray at a rate of 1.1 kg/ha (1 lb/acre) controlled 100 percent ofS. molesta in 1 week. Diquat at 4.5 kg/ha (4 lb/acre) controlled 99 percent of the plants, but such a high rate was considered cost prohibitive. Lower rates of either product were less effective (< 85-percent plant kill) and were considered unacceptable since the quantity of surviving plant material would lead to a recurrence of the weed problem. Thomas and Room (1986a) also reported that scientists in Zimbabwe successfiiUy demonstrated the efficacy of paraquat against S. molesta. As a result of these studies, paraquat was used from the late 1960's through the 1970's to control S. molesta in Kenya, Sri Lanka, Botswana, Australia, and Papua New Guinea (Mitchell 1979b; Miller and Pickering 1980; Thomas and Room 1986a; Oliver 1993).

Diatloff et al. (1979) investigated the use of kerosene plus surfactant (calcium dodecylbenzene sulfonate) mixed with and without the herbicides 2,4-D, dichlorprop, and diuron. Kerosene and surfactant were added to maximize herbicide penetration into the plant, since upper frond surfaces are covered with numerous trichomes or hairs, which can impede optimal herbicide coverage (Holm et al. 1977; Ohver 1993). Although kerosene plus surfactant alone inhibited S. molesta by 80 percent, the data showed that addition of herbicide significantly improved treatment performance. Rates as low as 0.15 kg/ha diuron

Chapters Management Options 15

added to the kerosene-surfeclant mixture controlled 100 percent of S. molesta. DicMorprop and 2,4-D controlled 92 to 95 percent of S. molesta in these trials, however higher rates of both products (0.75 to 3.00 kg/ha) were required. Because of the findings of these studies, a commercial product known as AF 101 (diuron mixed with kerosene and calcium dodecylbenzene sulfonate) was formulated and used successfully for large-scale treatment of S. molesta and other nuisance floating plants (e.g., Azollaflliculoides var. rubra and Pistia stratiotes) in Australia.

Scientists in New Zealand evaluated fluridone for herbicidal activitj' against S. molesta in outdoor tank studies (Wells et al. 1986). Both the granular and liquid formulations of fluridone were tested at concentrations ranging from 0.003 to 100 mg/L (ppm). Fluridone sjTnptoms (bleaching of new tissues) were noted on plants treated with concentrations >.0.1 mg/L fluridone, but plant death occurred only when plants were subjected to 100 mg/L fluridone. Both fluridone formulations performed similarly in these trials. In contrast to results by Wells et al. (1986), aquatic plant managers in Florida have recently reported successfiil control of S. molesta in small ponds with fluridone concentrations of 451» 90 |ig/L (ppb).' The rates used in Florida are an order of magnitude lower than those tested in New Zealand. The current maximum allowed label rate of application for fluridone in U.S. watere is 150 ng/L.

Hert}fclcie investigation and use in the United States

As S. molesta gained notoriety in the late 1990's as an established and noxious weed in the United States, concern for identifying management options escalated. While early studies conducted overseas identified several effective herbicides for S. molesta, only a few of these products were registered by the U.S. Environmental Protection Agency (USEPA) for use in aquatic environments. It was apparent that updated information on the use and efficacy of U.S.- labeled aquatic herbicides was needed.

Currently, eight herbicides are available in the United States for controlling weeds in aquatic habitats. They include: diquat, copper (as copper chelates), endothall, fluridone, gl}^hosate, imazapyr, 2,4-D, and triclopyr. Triclopyr and imazapyr are the most recent additions to this list, receiving Ml aquatic registra- tion from the USEPA in December 2002 and September 2003, respectively. To date, no information has been published on the effectiveness of triclopyr on S. molesta. Current information documenting the efficacy of the other available aquatic herbicides against S. molesta is summarized below.

Nelson et al. (2001) compared the response of six different herbicides (diquat, two formulations of endothall, glj'phosate, copper, and imazapyr) applied alone and in combination against S. molesta in an outdoor tank experiment. Tj'pe of surfactant, rate of application, and application technique were also examined. Of the 32 treatments evaluated, diquat (1.12 kg/ha) and glj^hosate (8.97 kg/ha) were equally effective for controlling S. molesta (99- to

Personal communication, J. D. Schardt, Florida Department of Environmental Protection, Tallahassee, FL.

16 Chapters Management Options

100-percent reduction in plant biomass). Copper (as the formulation KomeenTM) provided 81-percent control of 5'. molesta 42 days after treatment. Treatment with endothall (both formulations) showed significant plant control (80 to 86 percent) shortly after treatment, however, remaining plants were healthy and actively growing, resulting in a steady decline in percent control over time. Salvinia molesta was only minimally affected by the rates of imazapyr evaluated in this study.

Fairchild et al. (2002) determined the effectiveness of glyphosate mixed with several surfactants for control of .S. molesta. Five concentrations of glyphosate (0, 0.45, 0.91, 1.82, and 3.60 percent v:v) and five surfactants were evaluated under outdoor conditions. Results indicated that a 0.45-percent solution of glyphosate with or without surfactant significantly reduced plant biomass compared to untreated plants. Of the surfactants tested in this study, only glyphosate (0.45 percent v:v) mixed with Optima™ (0.25 percent v:v) resulted in complete mortality of plants with no regrowth. The results of this study demonstrated that S. molesta was sensitive to lower doses of glyphosate than was previously reported. Similar results with low-dose glyphosate treatments have been con- firmed in larger-scale outdoor experiments (unpublished data) by Dr. L. Nelson, a co-author of this report.

Glomski et al. (2003) investigated the use of the chelated copper formulation, Clearigate™, for control of 5. molesta. Results showed that Clearigate™ mixed as a 15- to 20-percent solution (v:v) was effective for reducing biomass by 84 to 88 percent. Regrowth of surviving plant tissues was observed indicating follow- up applications may be necessary to maintain acceptable containment. Although 100-percent plant mortality was not observed, the authors commented that the use of Clearigate™* may be a feasible altemative for managing S. molesta in areas where water use restrictions would prohibit the use of either diquat or glyphosate.

Several cases have been reported where operational-scale herbicide applications were successfiil in managing S. molesta infestations. State apphcators in Texas reported excellent control of iS. molesta using 4.54 kg/ha glyphosate (as RodeoTM) mixed with nonionic (Aqua-king™) and organo-silicone (Thorough- bredT!^ surfactants and applied in a total spray volume of 935 L water/ha (100 gal water/acre).''^ In 1995, officials in Soutii Carolina used diquat (2.24 to 3.36 kg/ha) and follow-up applications of fluridone (1.46 kg/ha) to eradicate S. molesta from a small plantation pond.^ As mentioned previously, fluridone has also been successfully used to control S. molesta populations in Florida. Most recentiy (March 2003), nearly 121 ha of 5. molesta were cleared from Lake Wilson, Oahu, Hawaii, using glyphosate (as AquaMaster™) applied at 4.6 to 5.7 kg/ha and mixed with a nonionic surfactant. Multiple applications were necessary in some areas where plants had produced a thick, vegetative mat. In

Personal communication, B. Kellum, San Jacinto River Authority, Conroe, TX. Personal communication, H. Elder, Texas Parks and Wildlife, Jasper, TX. Personal communication, S. de Kozlowski, South Carolina Department of Natural Resources,

Columbia, SC.


addition to herbicides, mechanical excavatore were also used at Lake Wilson to aid in the removal of plant biomass.'

Overall, the results of recent research and large-scale field applications have demonstrated that registered aquatic herbicides can be used to effectively manage S. molesta infestations in Ihe United States. While research on chemical management strategies for S. molesta is ongoing, to date, glj'phosate and diquat show the most promise. If used properly (correct rate, appHcation equipment, and application technique), either product can be expected to provide > 95 percent control of spraj^ed plants. Future research will continue to investigate new products and formulations, herbicide combinations, application techniques, and the potential for integrating management strategies.

Biological Control

Background

Use of one insect species, the weevil Cyrtobagous salviniae Calder and Sands (Figure 3), is recognized throughout the world as the method of choice for S. molesta management. Application of this agent has resulted in control in the tropical areas of 12 countries on 3 continents including, but not limited to, Australia, Fiji, India, Kenya, Namibia, South Afiica, Sri Lanka, Zambia, and Zimbabwe (Room et al. 1981; Fomo 1985; Fomo and Bourne 1985; Giliomee 1986; Joy et al. 1986; Thom^ and Room 1986b; Fomo 1987; JuMen and Griffiths 1998).

While never intentionally released, C. salviniae is found in the United States in many areas of southern Florida. Based on existing collection records, it was probably introduced accidentally from South America in the early 1960's. It is commonly found feeding on S. minima Baker, a close relative of S. molesta. Until recently, the Florida strain of C. salviniae was mistakenly identified as C. singularis, a species initially used for S. molesta control overseas but later proven to be noneffective. Attempts to use the Florida strain of C salviniae in the late 1990's and early 2000's to manage S. molesta in Texas and western Louisiana were not successfiil. Reasons for the failure are unknown but are believed related to conditioning of the Florida strain to S. minima for many years. This prompted researchere to petition the Technical Advisory Group (TAG) and the USD A, Animal and Plant Health Inspection Service (APHIS), Plant Protection and Quarantine (PPQ) for approval to release the Australian strain of C salviniae proven to be highly effective in oversea applications. Permission was granted in 2001, allowing the release of the Australian strain of C. salviniae in western Louisiana and east Texas only.

Personal communication, L. NaMiara, Hawaii Department of Agriculture, Honolulu, HI.

18 Chapters Management Options

Figure 3. Photograph of Cyrtobagous salviniae on S. minima (from USAGE 2001 and USDA, Agricultural Research Service (ARS))

Host-specificity testing

Before C. salviniae could be released, intensive host-specificity experiments were conducted. Various plants were offered to adults and immatures to docu- ment if feeding, oviposition, and survival could occur. Because of the extensive range of plants used by the Australians and South Africans in documenting host specificity, no additional testing was performed in the United States with approval for release based on the work of researchers fi-om both these countries (Fomo et al. 1983; Tipping and Center 2001).

Table 1 lists the plants used in host-specificity testing of C. salviniae in Australia and South Africa and for approval for release in the United States (Tipping and Center 2001). Based on these experiments, it was apparent that C. salviniae is highly host specific to S. molesta with only minor feeding occurring on Pistia strattotes L., another introduced and problem plant in the United States. Cyrtobagous salviniae fed extensively on all Salvinia spp. tested, which is not surprising. However, this is not a problem since native Salvinia do not occur in the continental United States or Hawaii.

Chapter 3 Management Options 19

Table 1 Plants Used In Host-S peciflcity Testing Species Name Common Name Family

Pteridophyta

SaMnia molesta Giant Salvinia Salvinias»a SaMnia minima Common Salvinia Saivinlacea Salvinia hastata Saiviniacea Azolla caroliniana Vteiter Fern Azoiiaraae Azolla tiliculoides Pacific mosquitofern Azolla ceae Azolla pinnata Africana Feathered mosquitofern Azollaceae Adiantum hispidulum Rougti Maidenhair Pteridaceae PteMum esculentum Braclcen Fern Denn^edtiaceae Marmlea dnimmondii Nardoo Marsileaceae Schizaea dichotoma Comb Fern Schizaeceae Christella dentate Binung Theiypteridaceae

Monocot^edons

Saglttaria granvnea An-ow Head Alismataceae Allium cepa Onion Amaryliidaceae Pistia stratiotes V\feteriettuce Arareae Ananas comosus Pineapple Bromeliaceae Zea mays Maize Poaceae Oryza sativa Rice Poaceae Sacchamm ofndnanim Sugarcane Poaceae Asparagus olTidnalls Asparagus Liliaceae Musa X para^fimacal Banana Musaceae Eichhomia crassjpes Wfaterhyacinth Pontederiaceae Potamogeton tricarinatus Floating Pondweed Potamogetonaceae Typha orientalls Bullrush Typhaceae Zingiber oMdnale Ginger Zinqiberaceae

II Dicotyledons

Carica papaya Papaya Caricaceae Beta vulgaris Beetroot Chenopodiaceae Spinada oleracea Spinadi Chenopodiaceae Lactuca sativa Lettuce Asteraceae Ipomoea batatas Sweet Potato Convolvulaceae Ipomoea aquattca Potato Vine Convolvuiaceae Cucurbita maxima Pumpkin Curcurbitaceae Rorippa nasturOum-aquaOcum \Afeter Cress Bra ssica ceae Brassca oleracea var. botrytis Cauliflower Brasacareae Medicago saliva Lucerne Fabaceae Trifolium subterraneum Subten^nean Clover Fabac»ae Gossypium hirsutum Cotton Malvaceae Nymphoides indica Wfeter Snowflal<e Menyanthaceae Eucalyptus tereOcomis Forest Redgum Myrtaceae Eucalyptus maculate Spotted Gum Myrtaceae Nymphaea gigantea Puiple V\tetertily Nymphaea ceae Ludwigia peploidBS V\bter Primro^ Onagraceae Polygonum iapathifolium Pale Knotweed Polygonareae Polygonum hydropiper V\feter Pepper Polygonaceae Polygonum sp. Smartweed Polygonaceae Rumex brownii Swamp Docic Polygonaceae Rumex crispus Curled Dock Polygonaceae Fragariaxananassa Strawberry Rosaceae Citrus mnenms Orange Ruteceae Citrus llmon Lemon Rutac»ae Citrus reOcuiata Mandarin Rutareae Lycopersicon esculentum Tomato Solanaceae

20 Chapter 3 Management Options

Description

Cyrtobagous salviniae is a small weevil ranging in length from 1.5 to 2.0 mm (USAGE 2001; Center et al. 1999). Adults are typically black, but newly emerged individuals may often be brown. Legs are reddish-brown in coloration. The dorsal surface of the weevil is covered with numerous shallow depressions or punctures as well as yellow peltate scales. The larvae are white and attain lengths of only about 3 mm. The pupa forms a cocoon in tiie modified subinerged fronds.

Biology

Adults typically reside on or beneath the fronds of S. molesta (USAGE 2001; Genter et al. 1999). A thin film of air adheres to the bottom of the weevil allowing for respiration during periods of submergence. Eggs are laid individually in cavities formed by the female's feeding activity. Egg hatch occurs in approximately 10 days. Total larval development requires 3 to 4 weeks. The prepupal and pupal periods last about 2 weeks.

Feeding damage

Adults will feed on the fronds, leaving small irregularly shaped holes, but prefer feeding on newly formed buds. Larvae feed within the fronds, rhizomes, and buds, and their feeding action can be devastating. Initial establishment is highly dependent on the nitrogen content of the plants. Australian scientists have found that applying nitrogen in the form of urea either directly or indirectly to the plants significantly increases the chance for establishment and initial population buildup (Fomo and Bourne 1985; Room and Thomas 1985).

Plant impacts

Feeding action of both the adults and larvae can be devastating with reported impact to field populations in other countries observed in several months instead of years as typically seen with other biological control agents (Room et al. 1981; Fomo 1987). During the early damaging phase, the plants initially turn brown in small patches that coalesce until finally the entire mat appears brown and begins to sink.

U.S. operational status

As indicated previously, permission was received in 2001 and C. salviniae weevils of the Australian strain were released shortly thereafter at sites in Texas and western Louisiana. While it is too early to confirm long-term establishment and impact, early evidence indicates that the weevils are estabUshed, expanding in distribution, and begiiming to impact Salvinia locally in the release areas.'

' Personal commimication, T. Center, U.S. Department of Agriculture, Agricultural Research Service, Aquatic Plant Management Laboratory, Fort Lauderdale, FL.


Recent controlled experimentation has indicated that previously observed differ- ences between the Florida and Australian weevil strains were erroneous, and this has prompted researchere to petition USDA, APHIS, and PPQ to accept bolli the Florida and Australian strains as the same.

Physical Control

Conventional procedures for floating aquatic plants

Physical alternatives for controlling free-floating aquatic plants have generally involved the following management practices: (a) physically removing the target species directly by hand, or by using water-based or land-based mechanical harvesters; (b) destroying tiie infestation by inflicting physical damage by cutting, shredding, or chopping targeted plants in situ; (c) placing barrier materials to confine nuisance vegetation to prevent colonization of other areas; and (d) modifying the environment to minimize gro%vth of the problem plant by eliminating or reducing suitable habitat (Chilton et al, 2002). Probably the least effective of these methods in controlling S. molesta infestations is the use of cutters, shreddere, or choppers that could spread Ae species by generating a large number of fiagmente (Madsen 2000; Chilton et al. 2002). These machines are also likely to kill nontarget biota and create water quality problems from decaying plant material. Pros and cons of each method are discussed in greater detail by Miller and Wilson (1989), Madsen (1997 and 2000), Wade (1990), and Chilton etal. (2002).

Direct removal

Manual removal has been successfijl in controlling S. molesta in the initial, imcrowded stage of population development (Thomas and Room 1986a; Miller and Wilson 1989). Once the plant has become established, hand removal and harvesting are less practical considering the plant's rapid rate of growth and production of large amounts of biom^s. Generally, the cost of equipment is expen- sive because the machinery is specialized to collect and remove large quantities of biomass in an aquatic environment (Thomas and Room 1986a; Chilton et al. 2002). With > 95-percent water content, S. molesta can be very heavy and can tj'pically double in weight in < 10 days under favorable conditions. In the tropics, its biomass has been recorded to reach approximately 200 tons (wet wt)/ha and in extreme cases, > 400 tons/ha within a 10-day period (Room and Julien 1994; Storrs and Mien 1996). In order to be effective, the harvester must be able to handle such biomass and remove it at rates exceeding regrowlh of the targeted vegetation (Storrs and Julien 1996). Even in winter in Australia, when S. molesta doubling rate was 40 to 60 days, the capacitj' of a large infestation for regrowth exceeded the removal capacitj' of machines (Mitchell 1979a). Furthermore, the morphometry of natural waterways may impose problems of inaccessibility and physical obstacles (e.g., other vegetation), to make large-scale mechanical removal impractical.


Barriers

Floating booms and nets have been used successfully to restrict migration of S. molesta downstream in rivers, streams, and other flowing systems (Farrell 1978; Finlayson and Mitchell 1982; Miller and Wilson 1989). They have also been deployed in very localized areas to prevent the plant from entering into and clogging water intakes, boat launches, marinas, and swimming areas (Chilton et al. 2002). However, booms and nets require continuous inspection and main- tenance and are subject to breakage under the pressure of large windblown mats (Oliver 1993). In some cases, a series of nets must be installed as protection against breakage of one or more nets during periods of flooding (Miller and Wilson 1989). In Lake Moondarra, netting supported on a boom of wire hawser, held afloat with 200-L drums, and anchored to concrete blocks was successful except in high-water flows when the boom was pushed free by the buildup of S. molesta (Finlayson and Mitchell 1982). Booms slung on 5-cm-diam steel cables in Lake Kariba were abandoned after being broken several times by windblown S. molesta accumulations (Thomas and Room 1986a).

Habitat alteration

Water-level drawdown is a relatively inexpensive technique for controlling aquatic weeds in lakes with sufficient water-level control structures (Chilton et al. 2002). The primary goal of the drawdown is to destroy the target plant by thorough drying and/or exposure to lethal (freezing) temperatures (Cooke et al. 1986). The success of this procedure in controlling S. molesta would largely depend on the structure of the plant mat and its ability to protect and insulate embedded ramets. Where the mat is thick and dense, the most exposed plants and plant parts are subject to desiccation and/or being frozen, while plants close to the sediment may survive. Unless the sediment becomes dry (and/or frozen) for a long enough period, plants deep in the mat may regenerate the colony on the return of suitable conditions.

Lantz et al. (1964), Lantz (1974) and Cooke et al. (1986) have reported water-level manipulation to be an important technique for controlling nuisance plants in many Louisiana reservoirs. Their findings indicate this procedure to be highly species specific and suggest that while drawdown may curtail one plant nuisance, it may promote development of a resistant species. In northern Louisiana, autumn/winter drawdown is most effective in reducing dense growth of £■. crassipes; however, the effectiveness of this technique on other plant species and in other locations may improve (or be lessened) by climatic conditions. In addition, water-level drawdown as a management tool has been some- what controversial because of adverse impacts on secondary uses of the aquatic system, e.g., boat access, hunting, fishing (Chilton et al. 2002). Recent evidence from pond studies in Lewisville, TX, has shown winter drawdown to be successful in reducing growth of S molesta (Dick et al., in review). Yet, further study is needed to determine efificacy of drawdown on S. molesta in other areas and when used in an integrated approach with other control methodologies.


Ecosystem Approach

Prevention

Obviously, prevention is the most effective, economical, and environ- mentally compatible method for dealing with an invasive plant like S. molesta. The spread of tiiis aquatic weed in the United Slates (and in other countries) has been closely Unked to the water garden trade where implementation of preventa- tive measures would be highly beneficial. Recently, S. molesta was found for sale in commercial nureeries in 12 states (Texas, Louisiana, Califomia, Hawaii, Arizona, New Mexico, Alabama, Oklahoma, Oregon, Washington, Virginia, and Pennsylvania) and in cultivation in many sites including botanical gardens, aquatic plant nurseries, and private ponds (Figure 2; Jacono 2003a). These activities are widely agreed to be principal mechanisms of plant spread, for example, when the plant is thrown by gardenere into nearby waters, or floods carry it from cultivation sites into natural drainages (Jacono 2003a). Water resources could be better protected by closer monitoring of public areas (e.g., boat launches, beaches, and marinas) and by educating users of the potential harm&l effects of S. molesta on aquatic systems and of the dangere of allowing this plant to escape cultivation. Additionally, State and Federal Agencies responsible for enforcing exotic species regulations (restricting the sale, transport, and cultivation of S. molesta) should be notified immediately of any possible violations and provided sufficient funding for inspection and enforce- ment pereonnel (Chilton et al. 2002).

Eradication

Barring prevention, early detection followed by prompt management action would help to eradicate S. molesta in its initial stages within a water body. However, given past experiences with this species and other invasive ^uatic plants, eradication, even within any single system, is almost always unattainable. It seems more likely that management approaches will need to be developed that seek to reduce the extent of 5. molesta infestations to acceptable levels.

Integrated pest management

When eradication is no longer feasible, integrated pest management (IPM) has proven effective in controlling weedy species, including E. crassipes. A balanced, IPM approach, emplojdng all appropriate control methods (chemical, biological, physical, and public outreach) will likely be required 1» "manage" the spread of S. molesta. This approach—^while it might include attempts at local eradication and would almost certainly include attempts at containment (prevention of spread)—^would seek to suppress the gro%^th or extent of S. molesta infestations, thereby reducing the plant's harmful enviroimiental or economic efifects.

The IPM approach utilizes knowledge of the biology and ecologj? of the target plant, knowledge of the efficacy and environmental effects of available control technologies, and continual monitoring of the infested system to allow


managing agencies to minimize the impacts of the offending plant. Ideally, this approach would rely heavily on the use of host-specific biocontrols to provide long-term, sustainable weed control.

"Filling the niche"

A key component of any long-term management strategy has to be sustain- ability. The most effective way to achieve sustainable control is by "fiUing the niche" with a beneficial, native species that will help to prevent a recurrence of the invasive species. Unfortunately, the niche occupied by S. molesta, a free- floating aquatic plant, is one that generally causes problems for managing agencies, whether filled by an exotic or native species. It is important, neverthe- less, to ensure the presence of a diverse community of aquatic plants as a defense against an overgrowth of an invasive species—whether submersed, free-floating, or emergent.

Nutrient management

Given the high rates of nutrient loading suffered by many aquatic systems, the development of any floating plant species potentially could be excessive. However, well-developed riparian wetland and littoral plant communities in lakes and reservoirs could intercept excess nutrients before they accumulate and promote the excessive growth of problematic species (Mitsch and Gosselink 1993). Since free-floating plants do not often have to compete with other plants for light, and because they depend on the water column for their nutrient supply, free- floating plants are more hkely to be limited by the availability of nutrients than are submersed and floating-leaved forms that obtain nutrients mainly from sediment (Sculthorpe 1985; Barko et al. 1991). Thus, reductions in nutrient supply are likely to provide corresponding reductions in biomass or growth rates of floating plants such as S. molesta.

Normally, waterbodies receive inputs of nutrients and sediment from the sur- rounding drainage basin as a result of natural runoff and soil erosion (Gupta 1979; Miller 1994). Problems with aquatic weeds, such as S. molesta, tend to arise when nutrient loadings are accelerated by human activities, a process known as cultural or anthropogenic eutrophication (Gupta 1979; Wetzel 1983). Nutrient inputs may become excessive because of agricultural runoff, failing septic systems, deforestation, building and road construction, and effluents from sewage treatment facilities (Miller 1994; Park 1997). Management strategies to reduce nutrient inputs through wise land-use and conservation practices (e.g., the use of advanced sewage treatment technologies, diversion of animal wastes, strip cropping and contour farming, and wetland protection) would help to improve water quality and reduce the potential for proliferation of 5'. molesta.


References

Ashton, P. J., and Mitchell, D. S. (1989). "Aquatic plants: Pattern and modes of invasion, attributes of invading species and assessment of control programmes." J. A. Drake, H. A. Mooney, F. di Castri, R. H. Groves, F. J. Krager, M. Rejmanek, and M. Williamson, eds.. Biological invasions: a global perspective. Wiley, New York, 111-154.

Barko, J. W., Gunnison, D., and Carpenter, S. R. (1991). "Sediment interactions with submersed macrophjle growlh and community dynamics" Aquatic Botany 41,41-65.

Barrett, S. C. H. (1989). "Waterweed invasions," Scientific American 260,90-97.

Bennett, F. D. (1975). "Insects and plant pathogens for the control oiSalvinia and Pistia."" Proceedings of the biological controlfijr water quality enhancement workshop. P. L. Brezonik and J. L. Fox, eds.. University of Florida, Gainesville, 28-35.

Bonnet, A. L. M. (1955). "Contribution a I'etude des hydroplerides: Recherches sar Salvinia auriculata Auhl," Annales des Sciences Naturelles Botanique et Biologic Vegetale 16, 529-601.

Boughey, A. S. (1963). "The explosive development of floating vegetation on Lake Kariba," Adansonia 3,49-61.

Cary, P. R, and Weerte, P. G. J. (1983a). "Growfli of Salvinia molesta as affected by water temperature and nutrition. I. Effects of nitrogen level and nitrogen cxympovmAs^' Aquatic Botany 16, 163-172.

. (1983b). "Growth of Salvinia molesta as affected by water temperature and nutrition. II. Effects of phosphorus levels" Aquatic Botany 17,61-70.

. (1984). "Growth of Salvinia molesta as affected by water temperature and nutrition. III. Nitrogen-phosphorus interactions and effect of pH/'Aquatic Botany 19, 171-182.

Center, T. D., Diay, F. A, Jubinsky, G. P., and Grodowitz, M. J. (1999). "Insects and other arthopods that feed on aquatic and wetland plants," USDA, Agricultural Research Service, Technical Bulletin 1870.

26 References

Chilton, E, Jacono, C. C, Grodowitz, M, and Dugas, C. (2002). ''Salvinia molesta: Status report and action plan," Unpublished report by the Giant Salvinia Interagency Task Force.

Coates, D. (1982). "Salvinia - possible biological effects on fish in Papua New Guinea?" Aquatics 4(3), 2.

Cooke, G. D., Welch, E. B., Peterson, S. A., and Newroth, P. R. (1986). Lake and reservoir management. Butterworth Publishers, Stoneham, MA.

Creagh, C. (1991/1992). "A marauding weed in check," Ecos 70,26-29.

Croxdale, J. G. (1978). "Salvinia leaves. I. Origin and early differentiation of floating and submerged leaves," Canadian Journal of Botany 56, 1982-1991.

(1979). "Salvinia leaves. II. Morphogenesis of the floating leaf," Canadian Journal of Botany 57, 1951-195 9.

_. (1981). "Salvinia leaves. III. Morphogenesis of the submerged leaf," Canadian Journal of Botany 59, 2065-2072.

Diatloff, G., Lee, A. N., and Anderson, T. M. (1979). "A new approach for salvinia control," Journal of Aquatic Plant Management 17,24-27.

Dick, G. O., Owens, C. S., and Smart, R. M. (in review). "Water level manipulation for controlling Salvinia molesta Mitchell in a Texas pond."

Divakaran, O., Arunachalam, M., and Nair, N. B. (1980). "Growth rates of Salvinia molesta Mitchell with special reference to salinity." Proceedings of the Indian Academy of Science. Plant Science 89, 161-168.

Fairchild, J. F., Allert, A. L., Riddle, J. S., and Gladwin, D. R. (2002). "Efficacy of glyphosate and five surfactants for controlling giant salvinia," Jowma/ of Aquatic Plant Management AO, 53-58.

Farrell, T. P. (1978). "The spread and control oiSalvinia molesta in Lake Moondarra, Mt. Isa, Queensland." Proceedings of the first conference of the council of Australian weed science societies. Melbourne, 179-188.

. (1979). "Control of Salvinia molesta and Hydrilla verticillata in Lake Moondarra, North-West Queensland," Mawagewe^f of Aquatic Weeds, Australian Water Resources Council, Department of Natural Resources, Canberra, 57-71.

Finlayson, C. M. (1984). "Growth rates oiSalvinia molesta in Lake Moondarra, Mount Isa, Australia," y4^Maftc Botany 18, 257-262.

Finlayson, C. M., Farrell, T. P., and Griffiths, D. J. (1982). "Treatment of sewage effluent using the water fem salvinia," Water Research Foundation of Australia, Report No. 57.

References 27

Finlayson, C. M., and Mitchell, D. S. (1982). "Management of salvinia(&/vj«ia molesta) in Australia," Australian Weeds 2, 71-76.

Fomo, I. W. (1983). ''Native distribution of the Salvinia auriculata complex and keys to species identification,"^^waftcfiotowy 17, 71-83.

. (1985). "How quickly can insects control salvinia in the tropics?" Proceedings of the 10th Asian-Pacific weed science society conference, Chiangmai, Thailand. 1,271-276.

. (1987). "Biological control of the floating fem Salvinia molesta in north-eastern Australia: Plant-herbivore interactions," Bulletin of Entomological Research 77, 9-17.

Fomo, I. W., and Bourne, A. S. (1985). "Feeding by adult Cyrtobagous salviniae on Salvinia molesta under different regimes of temperature and nitrogen content and the effects on plant growth," Entomophaga 30,279-286.

Fomo, I. W., and Harley, K. L. S. (1979). "Hie occurrence of Salvinia molesta in Bimil" Aquatic Botany 6,185-187.

Fomo, I. W., Sands, D. P. A., and Sexton, W. (1983). "Distribution, biology and host specificity of Cyrtobagous singularis Mustache (Coleoptera: Curculionidae) for the biological control of Salvinia molesta,^' Bulletin of Entomological Research 73, 85-95.

Gaudet, J. J. (1973). "Growth of a floating aquatic weed, Salvinia, under standaid conditions," Hydrobiologia 41,77-106.

Giliomee, J. (1986). "The biological control of Kariba weed in the East Caprivi," African Wildlife 40,189-195.

Global Compendium of Weeds (GCW). (2004). "Species search - Genera begirming with the letter S: Salvinia molesta D.Mitch." Web site at http://www.hear.org/gcw/index/html.

Glomski, L. M., Nelson, L. S., and Skogerboe, J. G. (2003). "Clearigate treatments for control of giant salvinia," Journal of Aquatic Plant Management 41, 127-129.

Gupta, O. P. (1979). Aquatic weeds: Their menace and control. Today and Tomorrow's Printers and Publishera, New Dehli, India.

Harley, K. L. S., and Mitchell, D. S. (1981). "The biology of Australian weeds. 6. Salvinia molesta D. S. Mitchell," Journal of the Australian Institute of Agricultural Science 47, 67-76.

Harper, L. M. (1986). "Management plan: Salvinia molesta Mitchell," Unpublished report. Advisory Services Division, Ministry of Agriculture and Fisheries, Hamilton, New Zealand.

28 References

Hassler, M., and Swale, B. (2002). "Family Salviniaceae, genus Salvinia: World species list." Website for Salvinia species of fern in the world; http://homepages.caverock.net.nz/~bi/fem/saIvinia.htm.

Hattingh, E. R. (1961). "Problem oiSalvinia auriculata Aubl. and associated aquatic weeds on Kariba Lake," Weed Research 1, 303-306.

Holm, L. G., Plucknett, D. L., Pancho, J. V., and Herberger, J. P. (1977). The world's worst weeds: Distribution and biology. The University Press of Hawaii, Honolulu.

Jacono, C, and Pitman, B. (2001). ''Salvinia molesta: Around the world in 70 years," Aquatic Nuisance Species Digest 4, 13-16.

Jacono, C. C. (2002). "News and notes on Salvinia molesta and Salvinia minima'" Website for the USGS; http://salvinia.er.usgs.gov/index.html. See section entitled "Time Series Map," at http://salvinia.er.usgs.gov/html/ sm progression.html.

. (2003a). "News and notes on Salvinia molesta and Salvinia minima'" Website for the USGS; http://salvinia.er.usgs.gov/index.html. See section on S. molesta cultivation, http://salvinia.er.usgs.gov/html/ cultivation.html.

_. (2003b). "The biology oiSalvinia sp." Website for the Cooperative Agricultural Pest Survey; http://www.ceris.purdue.edu/napis/pests/gs/ facts/bio.html.

Jacono, C. C, and Richerson, M. M. (2003). "News and notes on Salvinia molesta and Salvinia minimal Website for the USGS; http://salvinia.er.usgs.gov/index.html. See section on S. molesta distribution in the U.S., http://salvinia.er.usgs.gov/html/distribution map.html.

Johnson, D. (1995). "Giant salvinia found in South CaxoXma'" Aquatics 17, 22.

Johnson, D. S. (1967). "On the chemistry of freshwaters in southern Malaysia and Smgapore,"' Archiv fur Hydrobiologie 63, 477-496.

Joy, P. J., Satheesan, N. V., Lyla, K. R., and Joseph, D. (1986). "Successful biological control of the floating weed Salvinia molesta Mitchell using the weevil Cyrtobagous salviniae Calder and Sands in Kerala (India)." Proceedings of the 10^^ Asian-Pacific weed science society conference, Chiangmai, Thailand 2, 622-626.

Julien, M. H., and Boume, A. S. (1986). "Compensatory branching and changes in nitrogen content in the aquatic weed Salvinia molesta in response to disbudding," Oecologia 70, 250-257.

Julien, M. H., and Griffiths, M. W. (1998). Biological control of weeds: A world catalogue of agents and their target weeds. 4* ed., CABI Publishing, New York.

References 29

Kam-Wing, L., and Furtado, J. I. (1977). "The chemical control oiSalvinia molesta (Mitehell) and some related t»xicological sbxMes," Hydrobiologia 56,49-61.

Kay, S. H. (2002). 'Weed alert! Salvinia molesta escapes in NC," Website for the U.S. Geological Survey; http://salvinia.er.usgs.gov/. See page on North Carolina Alert; http://salvinia.er.usgs.gov/html/north carolina_alert.html.

Lantz, K. E. (1974). Natural and controlled water level fluctuation in a backwater lake and three Louisiana impoundments. Wildlife and Fisheries Conunission, Baton Rouge, LA.

Lantz, K. E., Davis, J. T., Hughes, J. S., and Shafer, H. E. (1964). "Water level fluctuation: Its effects on vegetation control and fish population management." ProceerfOTgi* of the Southeast Association Game and Fish Commission 18,483-494.

Loyal, D. S., and Grewal, R K. (1966). "Cytological study on sterility in Salvinia auriculata Aublet with a bearing on its reproductive mechanism," Cytologia 31,330-338.

M^sen, J. D. (1997). "Methods for management of nonindigenous aquatic plants." Assessment and Management of Plant Invasions. J. O. Luken and J. W. Thieret, eds.. Springer, New Yoric, 145-171.

. (2000). "Advantages and disadvantages of aquatic plant management techniques," Miscellaneous Paper, MP-00-1, U.S. Army Engineer Research and Development Center, Vicksburg, MS.

Miller, G. T. (1994). Living in the environment: Principles, connections, and solutions. W^sworth Publishing Company, Belmont, CA.

Miller, L L., and Pickering, S. E. (1980). "Sa/v/oia, a noxious weed," Department of Primary Production, Div. Ag. Stock, Northern Territory, Australia, Agnote Ref No. 80/35.

Miller, L L., and Wilson, C. G. (1989). "Management of salvinia in the Northem Territoiy," Journal of Aquatic Plant Management 27,40-46.

Mitchell, D. S. (1969). "The ecology of vascular hydrophytes on Lake Kariba," Hydrobiolo0a 34,448-464.

. (1970). "Autecologieal studies of Salvinia auriculata Aubl." Ph.D. diss.. University of London, England.

. (1972). "The kariba weed: Salvinia molesta,'" TTie British Fem Gazette 10,251-252.

_. (1979a). "The incidence and management of Salvinia molesta in Papua New Guinea." Office of Environmental and Conservation, and Department of Primary' Industry, Konedobu, Papua New Guinea.

30 References

Mitchell, D. S. (1979b). "Aquatic weeds in Papua New Guinea," Science in New Guinea 6(3), 154-160.

Mitchell, D. S., Petr, T., and Viner, A. B. (1980). "The water fern Salvinia molesta in the Sepik River, Papua New Guinea," Environmental Conservation 7, 115-122.

Mitchell, D. S., and Thomas, P. A. (1972). "Ecology of water weeds in the Neotropics." UNESCO Technical Papers in Hydrology, No. 12.

Mitchell, D. S., and Tur, N. M. (1975). "The rate of growth of Salvinia molesta {S. auriculata Auct.) in laboratory and natural conditions," Journal of Applied Ecology 12, 213-225.

Mitsch, W. J., and Gosselink, J. G. (1993). Wetlands. X"^ ed. Van Nostrand Reinhold, New York.

Nelson, B. (1984). "'Salvinia molesta Mitchell: Does it threaten Florida?" Aquatics 6, 6-8.

Nelson, L. S., Skogerboe, J. G., and Getsinger, K. D. (2001). "Herbicide evaluation against giant salvinia," Journal of Aquatic Plant Management 39, 48-53.

Oliver, J. D. (1993). "A review of the biology of giant salvinia {Salvinia molesta Mitchell)," Journal of Aquatic Plant Management 31, 227-231.

Owens, C. S., Smart, R. M., Honnell, D. R., and Dick, G. 0. (in preparation). "Effects of pH on growth of Salvinia molesta Mitchell."

Park, C. (1997). The environment: Principles and applications. Routledge, New York.

Plants National Database. (2003). "Plant profile for Salvinia molesta Mitchell," Plants Database Website; http://plants.usda.gov/.

Rani, V. U., and Bhambie, S. (1983). "A study of the growth of Salvinia molesta in relation to light and temperature," Aquatic Botany 17, 119-124.

Room, P. M. (1983). "Falling-apart as a lifestyle - the rhizome architecture and population growth of Salvinia molesta," Journal of Ecology 17, 349-365.

(1986a). ''Salvinia molesta - a floating weed and its biological control." The ecology of exotic animals and plants. R. L. Kitching, ed., John Wiley, Brisbane, Queensland, Australia, 164-186.

. (1986b). "Equations relating growth and uptake of nitrogen by Salvinia molesta to temperature and the availability of mtmgen," Aquatic Botany 24, 43-59.

References 31

Room, P. M. (1988). "Effects of temperature, nutrients and a beetle on branch architecture of the floating weed Salvinia molesta and simulations of biological control," Journal of Ecology 76, 826-848.

. (1990). "Ecology of a simple plant-herbivore system: Biological control of salvinia," Trends in Ecology and Evolution 5, 74-79.

Room, P. M., and Gill, J. Y. (1985). "The chemical environment of Salvinia molesta Mitchell: Ionic concentrations of infested waters,'" Aquatic Botany 23,127-135.

Room, P. M., Gunatilaka, G. A., Shivanathan, P., and Femando, I. V. S. (1989). "Control of Salvinia molesta in Sri Lanka by Cyrtobagous salviniae." Proceedings, 7th international symposium on biological control of weeds. E. S. Delfosse, ed., 285-290.

Room, P. M., Harley, K. L. S., Fomo, I. W., and Sands, D. P. A. (1981). "Successful biological control of the floating weed salvinia," Nature 294, 78-80.

Room, P. M., and Julien, M. H. (1994). "Population-biomass dynamics and the absence of-3/2 self-thinning in the clonal weed Salvinia molesta" Australian Journal of Ecology 19, l^-TiA.

Room, P. M., and Thomas, P. A. (1985). "Nitrogen and establishment of a beetle for biological control of the floating weed salvinia in Papua New Guinea," Journal of Applied Ecology 22,139-156.

. (1986a). "Nitrogen, phosphorus and potassium in Salvinia molesta Mitchell in the field: Effects of weather, insect damage, fertilizers and age," Aquatic Botany 24,213-232.

. (1986b). "Population growth of the floating weed Salvinia molesta: Field observations and a global model based on temperature and nitrogen," Journal of Applied Ecology 23, 1013-1028.

Sculthorpe, C. D. (1985). The biology of aquatic vascular plants. Riward Amold, London.

Storrs, M. J., and Julien, M. H. (1996). Salvinia: A handbook for the integrated control of Salvinia molesta in Kakadu National Park. Australian Nature Conservation Agency, Darwin.

TenBruggencate, J. (2003). "We've seen fish gasping for air' at Lake Wilson," Website for HonoluluAdvertiser: http://the.honoluluadvertiser.eom/article/2003/Jan/l 6/ln/ln 12a.html.

Thomas, P. A., and Room, P. M. (1986a). 'Taxonomy and control of Salvinia molesta,''Nature 320, 581-584.

32 References

Thomas, P.A., and Room, P. M. (1986b). "Successfiil control of the floating weed Salvinia molesta in Papua New Guinea: A useful biological invasion neutralizes a disastrous one," Environmental Conservation 13, 242-248.

Tipping, P. W., and Center, T. D. (2001). "Proposed field release of the South American weevil Cyrtobagous salviniae Clader and Sands (Coleoptera: Curculionidae) for control of the aquatic fern giant salvinia, Salvinia molesta Mitchell (Hydropteridales: Salviniaceae)." Release Petition to the Technical Advisory Group (TAG), June 2001.

University of Florida, Center for Aquatic and Invasive Plants. (2000). "Aquatic plants in Pen-and-ink: Supplement." IFAS Publication No. SP243. IFAS Publications, University of Florida, Gainesville, FL.

U.S. Army Corps of Engineers. (2001). "Noxious and Nuisance Plant Manage- ment Information System, CD-ROM, Version 5.0," U.S. Army Engineer District, Vicksburg, Vicksburg, MS.

U.S. National Arboretum. (2003). "USDA plant hardiness zone map (Web Version)," USDA Miscellaneous Publication No. 1475, January 1990.

Wade, P. M. (1990). "Physical control of aquatic weeds." yl^waftc weed: Ecology and management of nuisance aquatic vegetation. A. H. Pieterse and K. J. Murphy, eds., Oxford University Press, New York, 93-135.

Wells, R. D. S., Coffey, B. T., and Lauren, D. R. (1986). "Evaluation of fluridone for weed control in New Zealand," Journal of Aquatic Plant Management 24, 39-42.

Wetzel, R. G. (1983). Limnology. X"^ ed., Saunders College Publishing, New York.

Whiteman, J. B., and Room, P. M. (1991). "Temperatures lethal to Salvinia molesta MitcheW'' Aquatic Botany 40, 27-35.

References 33

REPORT DOCUMENTATION PAGE Form Approved

0MB No. 0704-0188 Public reporting burden for fhis collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid 0MB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.

1. REPORT DATE (DD-MM-YYYY) June 2004

2. REPORT TYPE Final report

3. DATES COVERED (From - To)

4. TITLE AND SUBTITLE

Salvinia molesta D. S. Mitchell (Giant Salvinia) in the United States: A Review of Species Ecology and Approaches to Management

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S)

D. G. McFarland, L. S. Nelson, M. J. Grodowitz, R. M. Smart, C. S. Owens

5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

U.S. Army Engineer Research and Development Center, Environmental Laboratory 3909 Halls Feny Road, Vicksburg, MS 39180-6199; U.S. Army Engineer Research and Development Center, Lewisville Aquatic Ecosystem Research Facihty 201 E. Jones Street, Lewisville, TX 75057

8. PERFORMING ORGANIZATION REPORT NUMBER

ERDC/EL SR-04-2

9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES)

U.S. Army Corps of Engineers Washington, DC 20314-1000

10. SPONSOR/MONITOR'S ACRONYM(S)

11. SPONSOR/MONITOR'S REPORT NUMBER(S)

12. DISTRIBUTION / AVAILABILITY STATEMENT

Approved for pubUc release; distribution is unhmited.

13. SUPPLEMENTARY NOTES

14. ABSTRACT

Over the past 70 years, the free-floating aquatic fern Salvinia molesta D. S. Mitchell (giant salvinia) has spread from its native range in Brazil to many tropical and subtropical regions. Though iimocuous within its native range, elsewhere this species is an aggressive menace that has had devastating ecological and socioeconomic impacts on aquatic systems in parts of Africa, Sri Lanka, India, Australia, New Gumea, and the Philippines. In the United States, the plant is established in waterways in at least 10 states (mainly in the south) and is expected to continue to expand in areas generally where Eichhornia crassipes (Mart.) Sohns (water hyacinth) persists. Listed as a Federal Noxious Weed since 1984, S. molesta is prohibited from importation to the United States and from transport across state lines. Dense mats of X molesta can suppress growth of native vegetation and degrade water quality, fish and wildlife habitat, and numerous other ecological values. Notably, massive infestations have occurred in the Swinney Marsh Complex, Texas, in the Lower Colorado River, Arizona/California, and in Lake Wilson and Enchanted Lake, Hawaii.

This report presents a review of available information on the growth, distribution, and ecology of S. molesta. Information is provided on the plant's taxonomic status, its field characteristics, phenology, and spread overseas and in the United States. Growth responses ofS. molesta in relation to environmental variables (e.g., temperature, nutrients, light, pH, conductivity) are emphasized as

(Continued) 15. SUBJECT TERMS

See reverse.

16. SECURITY CLASSIFICATION OF:

a. REPORT

UNCLASSIFIED

b. ABSTRACT

UNCLASSIFIED

c. THIS PAGE

UNCLASSIFIED

17. LIMITATION OF ABSTRACT

18. NUMBER OF PAGES

40

19a. NAME OF RESPONSIBLE PERSON

19b. TELEPHONE NUMBER (Include area code)

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std. 239.18

14. (Concluded)

are impacts of the species on the emironment and other aquatic organisms. Different technologies (i.e., physical, chemical, biological, and integrated) appUed to control S. molesta infestations are discussed along with information on the effectiveness of these procedures and their need for fiiilher study.

16. IConcluded)

African pyle Aquarium wateimoss Aquatic systems Australian azolla Control, management Disturbance Environmental fectoi^ Eutrophication Invasive species Kariba weed Water fern Water spangles

Aquatic Plant Control Research Program Salvinia molesta D ...

Documents

Transcript of Aquatic Plant Control Research Program Salvinia molesta D ...